Maze Solving Quadruped Robot : A Study Using Microbit XGO Robot V2

Abstract

Maze-solving has long been acknowledged as a fundamental robotics challenge, serving both a practical and symbolic assessment of a robot's intelligence and autonomy. Navigating through unfamiliar surroundings with obstacles has been used as a foundation for assessing fundamental robotic skills in both older mechanical systems and today's intelligent machines. This makes maze-solving more than just an academic exercise; it reflects real-world problems in fields like planetary exploration, autonomous cars, autonomous food delivery, warehouse automation, and search and rescue missions.

This thesis investigates autonomous navigation within a maze environment with a Micro:bit XGO Robot Kit V2 by evaluating various ranging sensors. Different time-of-flight (ToF) laser and ultrasonic modules were evaluated. The Micro:bit XGO Robot Kit V2 is a quadruped robot dog that serves as an educational platform with potential for advanced robotics applications.

The methodology involves testing performance through detailed case studies and an experimental framework. The thesis also evaluates the educational benefits of this project for bachelor’s-level students, emphasizing skills in programming, robotics, and sensor integration.

Recommendations for future alternative high-performance sensors, such as motion tracking sensors that utilize 3-axis accelerometers and gyroscopes, as well as replacing single-direction ranging sensors with more advanced 2D and 3D scanning LIDAR sensors.

This work contributes to the field of educational robotics by demonstrating how cost-effective tools can be utilized for complex tasks, providing a scalable model for teaching robotics principles.